Using radar-based quantitative precipitation data with coupled soil- and groundwater balance models for stream flow simulation in a karst area

Abstract A distributed soil water balance model, combined with a lumped karst groundwater model was developed, aiming to simulate stream discharge in karst catchments. Both modelling frameworks, which can be used separately, are presented. They were tested for the Lauchert catchment in the south-west of Germany, using radar based precipitation data, with the focus on simulating flood events. A soil water balance model with 100 m grid resolution was combined with groundwater model compartments on a subcatchment base for the catchment covering 450 km 2 . Results show good fits to observed runoff data and could be verified with soil-moisture and groundwater level observations. Specifically, spatially distributed precipitation data has proven to be very helpful for modelling local storm events, which are often not captured by station based weather data. The presented approach is a promising step toward an early warning system for floods in karst catchments.